Fall prediction alarm device, system and method based on gait analysis

Abstract

The invention provides a fall prediction alarm device, system and method based on gait analysis, the device, system and method are attached to feet, ankles, knee joints and the like so as to collect data such as action lines, muscle contraction, joint movement tracks and joint pressure, a module wirelessly uploads gait monitoring data to a background, gaits are analyzed by integrating the data on the background, and the fall prediction alarm device, system and method based on gait analysis are provided. The states of rehabilitation patient evaluation, fall prediction, fall prevention and the like are comprehensively evaluated. Fall prevention analysis can be carried out by fusing multi-dimensional gait analysis indexes, and a more accurate fall gait analysis result is provided for fall prevention analysis early warning. Fall judgment is carried out based on the gait features, the occurrence of false alarm of fall can be reduced, and the management level is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent medical care, and particularly to a fall prediction and alarm method based on gait analysis, a fall prediction and alarm device based on gait analysis, an electronic device, and a computer-readable storage medium. Background Art

[0002] The significance of fall prevention for rehabilitation patients lies in reducing the occurrence of fall events, reducing the risk of patient injury, and improving the rehabilitation efficiency and the quality of life of patients. Falls can not only cause physical injuries to patients, such as fractures, soft tissue injuries, etc., but also cause psychological fear and anxiety in patients, affecting their rehabilitation confidence and progress.

[0003] The fall prevention schemes adopted in the prior art are, for example, the following several schemes:

[0004] 1. Wearable devices: Use various sensors and alarm devices, such as fall detectors, gait analyzers, etc., to monitor the activity status of patients in real time. Once abnormal behavior or a fall is detected, an alarm is immediately issued.

[0005] 2. Medication management: Reasonably use medications to avoid situations that increase the risk of falls, such as dizziness and low blood pressure caused by medication side effects.

[0006] 3. Intelligent monitoring system: Utilize modern information technologies, such as the Internet of Things, big data analysis, etc., to build an intelligent monitoring system for continuous health monitoring and risk assessment of patients.

[0007] Although the above schemes can comprehensively prevent and reduce the occurrence of fall events in rehabilitation patients through different means and methods. However, in the application process, there are still the following application defects:

[0008] First of all, the above several fall prevention schemes are single-dimensional prevention schemes in the ontology technology. For example, wearable devices mainly rely on acceleration sensors or fall detectors in the wearable devices to collect, analyze, and judge sensing data. Generally, such sensing detection data has only one-dimensional data, such as acceleration data. Therefore, it is impossible to fuse multi-dimensional indicators for fall prevention analysis and provide a more accurate fall gait analysis result for fall prevention analysis and early warning.

[0009] Secondly, the existing fall prevention analysis systems lack applications based on gait feature analysis and calculation and judgment. Generally, it is the initial state judgment of the detection data. For example, it is judged whether the detected acceleration exceeds a threshold. If it exceeds, an alarm is issued. This judgment method lacks gait feature analysis and is therefore not accurate enough. Summary of the Invention

[0010] In order to solve the technical problems existing in the prior art, the present invention provides the following technical solutions:

[0011] On the one hand, a fall prediction and alarm method based on gait analysis is provided. This method is implemented by an electronic device and includes:

[0012] S1. Collect multi-dimensional gait data vectors G of the feet, ankles, and knee joints of a rehabilitation patient through electrode sensors respectively:

[0013] G = [g1, g2, g3,..., g n ,

[0014] where gn represents the gait vector data collected by the electrode sensor at the nth part, including: the line-of-action vector data of the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint;

[0015] S2. Perform feature processing on the multi-dimensional gait data vector G to extract the gait data vector features of each part;

[0016] S3. Based on the historical gait analysis database, predict and configure the proportion weights of the gait data vector features of each part;

[0017] S4. Based on the gait data vector features and the proportion weights, construct a gait feature abnormality index and a fall risk index;

[0018] S5. Judge whether gait abnormality occurs according to the gait feature abnormality index:

[0019] If it occurs, judge the corresponding risk level according to the fall risk index, and match and output the fall alarm strategy corresponding to the risk level;

[0020] If not, repeat the above steps.

[0021] Preferably, S2. Perform feature processing on the multi-dimensional gait data vector G to extract the gait data vector features of each part, including:

[0022] Analyze the multi-dimensional gait data vector G to obtain the line-of-action vector data of the corresponding foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint respectively;

[0023] Extract the gait data vector features of the line-of-action vector data, the joint movement trajectory vector data, and the muscle contraction vector data respectively: F force , F joint , F muscle , where:

[0024]

[0025] where:

[0026] f i represents the foot pressure value collected by the i-th pressure electrode patch sensor, represents the average foot pressure, and k represents the number of pressure patch sensors;

[0027]

[0028] where:

[0029] θ(t) represents the joint angle collected by the ankle angle electrode patch sensor that changes with time t, and θ ref (t) represents the reference joint angle (the angle under normal gait), and T represents the gait cycle;

[0030] F muscle = max(EMG(t)) - min(EMG(t)),

[0031] where:

[0032] EMG(t) represents the muscle electrical signal that changes with time t.

[0033] Preferably, S3. Based on the historical gait analysis database, predict and configure the proportion weights of the gait data vector features of each part, including:

[0034] Statistically analyze the historical gait data vector features of each part and mark the corresponding abnormal labels;

[0035] Calculate the feature importance score I i of the historical gait data vector features of each part, where i represents the i-th feature;

[0036] According to the feature importance score I i , calculate the weight w i of the historical gait data vector features of each part:

[0037]

[0038] where n represents the total number of features, and I j represents the feature importance score of the j-th feature;

[0039] Construct a historical gait analysis database composed of several groups of {historical gait data vector features, weight w i}, and use each group of data in the historical gait analysis database as the training input of the RF random forest model for model training and learning to construct a weight prediction model;

[0040] Verify the model performance of the weight prediction model. If the verification is passed, deploy and apply the weight prediction model;

[0041] Predict and output the proportion weight w of the gait data vector features of each part by using the weight prediction model i .

[0042] Preferably, S4. Based on the gait data vector features and the proportion weight, construct a gait feature abnormality index and a fall risk index, including:

[0043] Construct a gait feature abnormality index S abnormal :

[0044] S abnormal = w1·F force + w2·F joint + w3·F muscle ,

[0045] wherein:

[0046] w1, w2, and w3 are the weights of the line of action, joint movement trajectory, and muscle contraction characteristics respectively, calculated according to the weight w i ;

[0047] When S abnormal > the threshold A, it is determined that gait abnormality occurs;

[0048] Construct a fall risk index R fall :

[0049]

[0050] wherein:

[0051] α and β are the weights of the change rates of the line of action and joint movement trajectory respectively,

[0052] When R fall > the threshold B1, it is determined to be a high fall risk, and there is a corresponding fall alarm strategy at this time;

[0053] When the threshold B1 > R fall > the threshold B2, it is determined to be a medium fall risk, and there is a corresponding fall alarm strategy at this time;

[0054] When the threshold B2 > R fall at this time, it is determined to be a low fall risk, and no fall alarm strategy is output.

[0055] On the other hand, a fall prediction and alarm device based on gait analysis is provided. The fall prediction and alarm device based on gait analysis is used to implement the above-mentioned fall prediction and alarm method based on gait analysis. The device includes:

[0056] A gait analysis terminal, including:

[0057] MCU;

[0058] An electrode patch sensor for respectively collecting multi-dimensional gait data vectors G of the foot, ankle and knee joint parts of a rehabilitation patient;

[0059] A signal processing module for performing feature processing on the multi-dimensional gait data vector G to extract gait data vector features of each part;

[0060] A wireless transmission module for uploading the gait data vector features to a background server;

[0061] A power supply for power supply;

[0062] A background server for predicting and configuring the proportion weights of the gait data vector features of each part based on a historical gait analysis database; and, constructing a gait feature abnormality index and a fall risk index based on the gait data vector features and the proportion weights; judging whether gait abnormality occurs according to the gait feature abnormality index: if it occurs, judging the corresponding risk level according to the fall risk index, matching and outputting a fall alarm strategy corresponding to the risk level; if not, repeating the above steps;

[0063] The electrode patch sensor, the signal processing module, the wireless transmission module and the power supply are respectively electrically connected to the MCU;

[0064] The gait analysis terminal is communicatively connected to the background server.

[0065] On the other hand, an electronic device is provided, which includes: a processor; a memory, on which computer-readable instructions are stored, and when the computer-readable instructions are executed by the processor, any one of the methods in the above-mentioned fall prediction and alarm method based on gait analysis is implemented.

[0066] On the other hand, a computer-readable storage medium is provided, in which at least one instruction is stored, and the at least one instruction is loaded and executed by a processor to implement any one of the methods in the above-mentioned fall prediction and alarm method based on gait analysis.

[0067] The beneficial effects brought by the technical solutions provided in the embodiments of the present invention at least include:

[0068] The present invention collects data such as locomotion force lines, muscle contractions, joint movement trajectories, and joint pressures by attaching patches to the feet, ankles, knees, etc. The module uploads gait monitoring data wirelessly to the background, where the data is comprehensively analyzed to evaluate gait and comprehensively assess the status of rehabilitation patients, fall prediction, fall prevention, etc. It can integrate multi-dimensional gait analysis indicators for fall prevention analysis, providing more accurate fall gait analysis results for fall prevention analysis and warning. Based on gait characteristics for fall judgment, it can reduce the occurrence of false fall alarms and optimize the management level. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0070] Figure 1 is a flowchart of a fall prediction and alarm method based on gait analysis provided by an embodiment of the present invention;

[0071] Figure 2 is a block diagram of a fall prediction and alarm device based on gait analysis provided by an embodiment of the present invention;

[0072] Figure 3 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0073] The following describes the technical solutions in the present invention with reference to the drawings.

[0074] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of the word "exemplary" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0075] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when their differences are not emphasized, their intended meanings are the same. "(of)", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when their differences are not emphasized, their intended meanings are the same.

[0076] In the embodiments of the present invention, sometimes subscripts such as W1 may be miswritten as non-subscript forms such as W1. When the difference is not emphasized, their intended meanings are the same.

[0077] To make the technical problems, technical solutions, and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0078] Embodiments of the present invention provide a fall prediction and alarm method based on gait analysis. This method can be implemented by an electronic device, which can be a terminal or a server. As Figure 1 shown in the flowchart of the fall prediction and alarm method based on gait analysis, the processing flow of this method can include the following steps:

[0079] S1. Through electrode patch sensors, collect multi-dimensional gait data vectors G of the feet, ankles, and knee joint parts of rehabilitation patients respectively:

[0080] G = [g1, g2, g3,..., g n ,

[0081] where gn represents the gait vector data collected by the electrode patch sensor at the nth part, including: the motive force line vector data of the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint part;

[0082] S2. Perform feature processing on the multi-dimensional gait data vector G to extract the gait data vector features of each part;

[0083] S3. Based on the historical gait analysis database, predict and configure the proportion weights of the gait data vector features of each part;

[0084] S4. Based on the gait data vector features and the proportion weights, construct a gait feature abnormality index and a fall risk index;

[0085] S5. Determine whether gait abnormalities occur according to the gait feature abnormality index:

[0086] If so, determine the corresponding risk level according to the fall risk index, and match and output a fall alarm strategy corresponding to the risk level;

[0087] If not, repeat the above steps.

[0088] The present invention provides a module system based on electrode patch sensing monitoring. It is pasted on the feet, ankles, knee joints, etc. to collect data such as motive force lines, muscle contractions, joint movement trajectories, and joint pressures. The module uploads gait monitoring data wirelessly to the background, and comprehensively analyzes the gait based on the data on the background to comprehensively evaluate the states of rehabilitation patients, fall prediction, fall prevention, etc.

[0089] Combined with the attached Figure 2 Shown application device / system, the present invention collects gait signal data of various parts through electrode patches, combines background gait analysis and algorithm calculation and judgment, so as to analyze the gait characteristics of rehabilitation patients and give corresponding fall warning strategies (different levels of alarm strategies and plans, which can issue alarms according to the administrator's settings and can send strategies for this patient from the background to the corresponding nurse PDA).

[0090] 1. Hardware part:

[0091] Electrode patch sensors: Attached to parts such as the foot, ankle, and knee joint to collect electromyogram signals (EMG), pressure signals, joint angle signals, etc.

[0092] Signal processing module: Filters, amplifies, and digitizes the collected original signals.

[0093] Wireless transmission module: Uploads the processed data to the background through Bluetooth or Wi-Fi.

[0094] Power supply module: Powered by a micro battery, supporting low-power operation.

[0095] The electrode patch sensors for each part are selected according to the requirements and installed on the corresponding parts after debugging, and then activated and put into use.

[0096] The specific principles and implementation methods for the electrode patch sensors to be attached to parts such as the foot, ankle, and knee joint to collect electromyogram signals (EMG), pressure signals, and joint angle signals are as follows:

[0097] 1). Collection of electromyogram signals (EMG)

[0098] Principle: Muscles generate electrical activity changes during movement. When a muscle is stimulated by a nerve signal, a potential difference is generated, thereby recording the current activity. The electrode patch sensor captures these electrical signals through electrodes attached to the body surface and transmits them to the connected electronic device for further amplification and display, and records them in the form of a graph to form an electromyogram.

[0099] Implementation method: The electrode patch is usually made of a conductive material, such as silver / silver chloride (Ag / AgCl), and is attached with an adhesive to ensure close contact with the skin. During the collection process, the electrode patch is attached to the target muscle group, such as the muscles on the sole of the foot, the muscles around the ankle joint, and the muscles in front of and behind the knee joint. Connect the electrode patch to an electronic device such as an electromyograph through a wire, and then the collection and analysis of electromyogram signals can be started.

[0100] 2). Collection of pressure signals

[0101] Principle: The pressure sensor can sense the pressure signal and convert it into an available output electrical signal. When a person is walking, running or performing other activities, different degrees of pressure will be exerted on parts such as the feet, ankles and knees. These pressure changes can be captured and quantified by the pressure sensor.

[0102] Implementation method: The pressure sensor is usually embedded in the electrode sheet in the form of a thin film or gasket, or attached to the target part as an independent sensor. During the acquisition process, the sensor converts the sensed pressure into an electrical signal and transmits it through a wire to the data acquisition system for analysis.

[0103] 3) Acquisition of joint angle signals

[0104] Principle: The joint angle sensor captures motion information by measuring the angle changes of the joint during movement. These sensors usually work based on resistive, inductive or optical principles and can accurately record the angle values of the joint in different states.

[0105] Implementation method: The joint angle sensor can be fixed to the joint part, such as the ankle joint or knee joint, through a rigid bracket or a flexible band. During the acquisition process, the sensor will record the angle changes of the joint in real time and transmit the data to the data acquisition system for analysis through wireless or wired means. In order to achieve more comprehensive motion analysis, the joint angle sensor can be used in combination with electromyogram signals and pressure sensors to obtain richer motion information.

[0106] 2. Software part:

[0107] Data acquisition and preprocessing: Denoise and normalize the original data, gait feature extraction: Extract gait features based on the vector data processing formula, data analysis and early warning: Combine the gait feature calculation and judgment formula to achieve rehabilitation evaluation, fall prediction and prevention.

[0108] The present invention integrates multi-dimensional gait analysis indicators for fall prevention analysis, providing a more accurate fall gait analysis result for fall prevention analysis and early warning. Making a fall judgment based on gait features can reduce the occurrence of false fall alarms and optimize the management level.

[0109] Preferably, S2. Perform feature processing on the multi-dimensional gait data vector G to extract the gait data vector features of each part, including:

[0110] Analyze the multi-dimensional gait data vector G to respectively obtain the action force line vector data corresponding to the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint part;

[0111] Extract the gait data vector features of the action force line vector data, the joint movement trajectory vector data, and the muscle contraction vector data respectively: F force , F joint , F muscle , where:

[0112]

[0113] where:

[0114] f i represents the foot pressure value collected by the i-th pressure electrode sensor, represents the average foot pressure, and k represents the number of pressure patch sensors;

[0115]

[0116] where:

[0117] θ(t) represents the joint angle that changes with time t collected by the ankle angle electrode sensor, and θ ref (t) represents the reference joint angle (the angle under normal gait), and T represents the gait cycle;

[0118] F muscle = max(EMG(t)) - min(EMG(t)),

[0119] where:

[0120] EMG(t) represents the muscle electrical signal that changes with time t.

[0121] The gait data can be represented as a multi-dimensional vector, and each dimension corresponds to a type of sensor data (such as joint angle, pressure distribution, muscle electrical signal, etc.). Each dimension corresponds to the action force line vector data of the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint, which can be described in combination with the above G and represented by the data collected by the corresponding electrode sensors.

[0122] In gait analysis, the action force line vector data of the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint respectively characterize the movement characteristics and functional states of the foot, ankle joint, and knee joint during the gait cycle. The following are the specific characteristics and meanings of these vector data:

[0123] I. Action force line vector data of the foot

[0124] 1. Feature description

[0125] The action force line vector data describes the magnitude and direction of the force on the foot during the gait cycle. Usually includes:

[0126] - Pressure distribution: The pressure values in each area of the sole of the foot.

[0127] - Line of force direction: The direction vector of the force on the foot.

[0128] 2. Characteristic features

[0129] - Foot stability: Through the pressure distribution and the line of force direction, the stability of the foot during gait can be judged. For example, a uniform pressure distribution and a stable line of force direction indicate good foot support.

[0130] - Gait symmetry: By comparing the pressure distribution and the line of force direction of the left and right feet, the symmetry of gait can be evaluated.

[0131] - Foot functional status: Abnormal pressure distribution or line of force direction may indicate foot diseases (such as flat feet, high arches) or injuries.

[0132] 3. Significance

[0133] - Gait analysis: Used to evaluate the functional status of the foot during gait and identify abnormal gaits.

[0134] - Rehabilitation assessment: Used for the foot function assessment of rehabilitation patients to guide rehabilitation training.

[0135] - Fall warning: By analyzing the change of the force on the foot, the fall risk can be predicted.

[0136] II. Vector data of joint movement trajectories of the ankle

[0137] 1. Feature description

[0138] The vector data of joint movement trajectories describes the movement trajectory of the ankle joint during the gait cycle, usually including:

[0139] - Angle change: The flexion, extension, inversion, and eversion angles of the ankle joint.

[0140] - Angular velocity: The angular velocity of the movement of the ankle joint.

[0141] 2. Characteristic features

[0142] - Ankle joint flexibility: Through the angle change and the angular velocity, the flexibility and range of motion of the ankle joint can be evaluated.

[0143] - Gait coordination: The movement trajectory of the ankle joint reflects the coordination of gait. For example, the flexion and extension angles of the ankle joint should be synchronized with the gait cycle.

[0144] - Ankle joint stability: Abnormal movement trajectories may indicate ankle joint instability or injury.

[0145] 3. Significance

[0146] - Gait analysis: Used to evaluate the functional status of the ankle joint during gait and identify abnormal gaits.

[0147] - Sports performance optimization: Used for the ankle joint function evaluation of athletes to optimize sports performance.

[0148] - Rehabilitation assessment: Used for the ankle joint function evaluation of rehabilitation patients to guide rehabilitation training.

[0149] III. Muscle contraction vector data of the knee joint

[0150] 1. Feature description

[0151] The muscle contraction vector data describes the contraction state of the muscles around the knee joint and usually includes:

[0152] - Electromyogram (EMG): The intensity of the electrical signals of the muscles around the knee joint.

[0153] - Muscle contraction force: The magnitude of the contraction force of the muscles around the knee joint.

[0154] 2. Characteristic features

[0155] - Muscle activity intensity: Through the electromyogram and contraction force, the activity intensity of the muscles around the knee joint can be evaluated.

[0156] - Knee joint stability: The muscle contraction state reflects the stability of the knee joint. For example, insufficient muscle contraction force may lead to knee joint instability.

[0157] - Gait coordination: The contraction state of the muscles around the knee joint should be synchronized with the gait cycle, and abnormal muscle contractions may indicate gait incoordination.

[0158] 3. Significance

[0159] - Gait analysis: Used to evaluate the functional status of the knee joint during gait and identify abnormal gaits.

[0160] - Sports performance optimization: Used for the knee joint function evaluation of athletes to optimize sports performance.

[0161] - Rehabilitation assessment: Used for the knee joint function evaluation of rehabilitation patients to guide rehabilitation training.

[0162] IV. Comprehensive application

[0163] 1. Gait analysis

[0164] By comprehensively analyzing the vector data of the foot, ankle joint, and knee joint, the functional status of the gait can be comprehensively evaluated and abnormal gaits can be identified.

[0165] 2. Rehabilitation assessment

[0166] Used to assess the foot, ankle and knee function of rehabilitation patients and to guide rehabilitation training.

[0167] 3. Optimize sports performance

[0168] Used to assess athletes' foot, ankle and knee function and optimize sports performance.

[0169] 4. Fall warning

[0170] Predict fall risk by analyzing changes in foot, ankle, and knee joint vector data.

[0171] Therefore, the action force line vector data of the foot, the joint activity trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint respectively characterize the motion characteristics and functional status of the foot, ankle, and knee joints during the gait cycle. These vector data have important application value in gait analysis, rehabilitation assessment, sports performance optimization, and fall warning. By comprehensively analyzing these vector data, the functional status of gait can be comprehensively evaluated, abnormal gait can be identified, and corresponding intervention measures can be taken.

[0172] Through the above multi-dimensional gait data vector features: F force , F jiont , F muscle The extraction and calculation of gait feature data of rehabilitation patients at different time points can be obtained, and further used to analyze gait, comprehensively evaluate the status of rehabilitation patients, fall prediction, and fall prevention.

[0173] Preferably, S3, based on the historical gait analysis database, predicting and configuring the weight of the gait data vector features of each part, includes:

[0174] Count the historical gait data vector features of each part and add corresponding abnormal labels;

[0175] Calculate the feature importance score I of the historical gait data vector features of each part i , where i represents the i-th feature;

[0176] According to the feature importance score I i , calculate the weight w of the historical gait data vector feature of each part i :

[0177]

[0178] Where n represents the total number of features, I j represents the feature importance score of the jth feature;

[0179] Construct several groups of {historical gait data vector features of each part, weight w i} to form a historical gait analysis database, and use each group of data in the historical gait analysis database as the training input of the RF random forest model for model training and learning to construct a weight prediction model;

[0180] Verify the model performance of the weight prediction model. If the verification passes, deploy and apply the weight prediction model;

[0181] Use the weight prediction model to predict and output the proportion weight w of the gait data vector features of each part i 。

[0182] Because in the process of gait abnormality analysis, the action force line, joint movement trajectory, and muscle contraction characteristics will each occupy a certain importance. For example, the joint characteristics of patients play a relatively large role in fall warning. At this time, the gait data vector characteristics of the joint movement trajectory vector characteristics and muscle contraction vector data will occupy an important position in the entire gait abnormality analysis. Therefore, it is necessary to consider the proportion weights of each feature in the gait feature abnormality analysis to accurately judge the fall risk level or status of the patient.

[0183] In the present invention, in order to quickly configure the weights of each feature, based on the historical gait analysis database, predict and configure the proportion weights of the gait data vector features of each part, and be able to automatically predict and allocate the weights of each feature according to the historical gait data of the user, realizing intelligent automation, flexibly adjusting and configuring the weights, and reducing the risk of human intervention.

[0184] Specifically: This solution is based on the historical gait analysis database, and through data analysis and machine learning methods, predicts and configures the proportion weights of each gait data vector feature. Through the statistical analysis of historical data and model training, determine the contribution degree of each feature to gait analysis and fall warning, so as to optimize the weight configuration and improve the accuracy and robustness of the algorithm.

[0185] 1. Data Preparation

[0186] Historical gait database: Contains gait data of a large number of users, such as walking speed, stride length, joint angle change rate, pressure distribution uniformity, etc.

[0187] Label data: Each gait data corresponds to a label (such as normal gait, abnormal gait, fall event, etc.)

[0188] 2. Feature Weight Configuration Method

[0189] Determine the weights of each feature through the following steps:

[0190] 1. Feature importance analysis: Use statistical methods or machine learning models to analyze the contribution degree of each feature to gait classification and fall warning.

[0191] 2. Weight assignment: Assign weights to each feature according to its importance.

[0192] 3. Weight optimization: Optimize the weight configuration through model training and validation.

[0193] III. Specific implementation steps

[0194] 1. Feature importance analysis

[0195] Use models such as Random Forest or Gradient Boosting Tree to calculate the importance scores of each feature. Let the feature importance score be, where i represents the i-th feature.

[0196] 2. Weight assignment

[0197] According to the feature importance score I i , calculate the weight w of the historical gait data vector features of each part i :

[0198]

[0199] where n represents the total number of features, and I j represents the feature importance score of the j-th feature.

[0200] Through the above processing, a historical gait analysis database composed of several groups of {historical gait data vector features, weight w i} of each part can be constructed based on the user's historical gait data, and each group of data in the historical gait analysis database is used as the training input of the RF random forest model for model training and learning to construct a weight prediction model.

[0201] Optimize the weight configuration through model training and validation. The specific steps are as follows

[0202] 1. Model training: Use historical gait data to train a classification model (RF random forest model);

[0203] 2. Weight adjustment: Adjust the weights of each feature according to the model performance (such as accuracy, recall, etc.);

[0204] 3. Validation and iteration: Test the model performance on the validation set and iteratively optimize the weight configuration.

[0205] The training and application of the RF random forest model will not be elaborated here.

[0206] For model validation, the administrator can reconstruct the validation set (constructed based on a similar principle to the training data) and verify the model performance based on, for example, recall.

[0207] and the weights w of each part are predicted i After that, the weight configuration can be optimized through model training and validation: w′ i :

[0208] w′ i = w i + α·Δw i ,

[0209] where α represents the learning rate, and Δw i represents the adjustment amount of the weight, which is calculated according to the model performance.

[0210] Preferably, S4. Based on the gait data vector features and the proportion weights, a gait feature anomaly index and a fall risk index are constructed, including:

[0211] Construct a gait feature anomaly index S abnormal :

[0212] S abnormal = w1·F force + w2·F joint + w3·F muscle ,

[0213] where:

[0214] w1, w2, and w3 are the weights of the line of action, joint movement trajectory, and muscle contraction characteristics respectively, which are calculated according to the weight w i ;

[0215] When S abnormal > the threshold A, it is determined that gait abnormality occurs;

[0216] Construct a fall risk index R fall :

[0217]

[0218] where:

[0219] α and β are the weights of the change rates of the line of action and joint movement trajectory respectively, which can be set by default in the system. The change rates of the line of action and joint movement trajectory are represented by the absolute value of the differential, which can comprehensively determine the fall risk by the foot force and joint angle change rate (because the muscle signal change has little significance in falling, and the patient's falling situation is mainly caused by the foot force and joint angle);

[0220] When R fall > the threshold B1, it is determined as a high fall risk, and there is a corresponding fall alarm strategy at this time;

[0221] When the threshold B1 > R fallWhen it is greater than the threshold B2, it is determined as medium fall risk, and there is a corresponding fall alarm strategy at this time;

[0222] When the threshold B2 > R fall When it is, it is determined as low fall risk, and no fall alarm strategy is output.

[0223] Through the above index, gait abnormality can be judged, and then fall risk can be judged.

[0224] The above threshold can be determined by the administrator according to the patient's condition. For example, if the patient can move independently, the corresponding threshold can be larger; on the contrary, for patients who need assisted walking, the threshold can be set smaller.

[0225] The background generates the corresponding risk level according to the judgment result, matches the fall alarm strategy corresponding to the risk level from the database, and sends it to the nurse PDA or the family member APP that manages the patient.

[0226] The fall alarm strategies corresponding to specific different risk levels are formulated by the administrator.

[0227] Figure 2 It is a block diagram of a fall prediction and alarm device based on gait analysis shown according to an exemplary embodiment. The device is used for a fall prediction and alarm method based on gait analysis. Refer to Figure 2 , the device includes:

[0228] The gait analysis terminal includes:

[0229] MCU;

[0230] Electrode sheet sensors, used to collect multi-dimensional gait data vectors G of the feet, ankles, and knee joint parts of rehabilitation patients respectively;

[0231] The signal processing module is used to perform feature processing on the multi-dimensional gait data vector G and extract the gait data vector features of each part;

[0232] The wireless transmission module is used to upload the gait data vector features to the background server;

[0233] Power supply, used for power supply;

[0234] The background server is used to predict and configure the proportion weights of the gait data vector features of each part based on the historical gait analysis database; and, based on the gait data vector features and the proportion weights, construct a gait feature abnormality index and a fall risk index; judge whether gait abnormality occurs according to the gait feature abnormality index: if it occurs, judge the corresponding risk level according to the fall risk index, match and output the fall alarm strategy corresponding to the risk level; if it does not occur, repeat the above steps;

[0235] The electrode sheet sensor, the signal processing module, the wireless transmission module and the power supply are respectively electrically connected to the MCU;

[0236] The gait analysis terminal is communicatively connected to the background server.

[0237] The gait analysis terminal can be designed in a style and principle similar to that of a wristband or a smart bracelet, on which patch electrodes for the foot, ankle and knee joints, a signal processor, an MCU, a Bluetooth wireless module, etc. are integrated, and it can process data based on the principles of the above method steps, so as to collect gait data and upload it to the background, and further perform gait early warning analysis on the background.

[0238] The overall steps are divided into:

[0239] Data collection and upload:

[0240] The electrode sheet sensor real-time collects gait data and uploads it to the background through the wireless transmission module.

[0241] Gait analysis and evaluation:

[0242] The background calculates gait features based on the vector data processing formula and the feature extraction formula, and combines historical data and reference values to evaluate the gait state of the rehabilitated patients.

[0243] Fall prediction and warning

[0244] Based on the fall risk index R fall , real-time monitor gait changes.

[0245] When a high fall risk is detected, trigger an early warning mechanism (such as sending an alarm to medical staff or family members).

[0246] Rehabilitation suggestions and interventions:

[0247] According to the gait abnormality index S abnormal , generate personalized rehabilitation suggestions, and design a rehabilitation training plan in combination with the joint movement trajectory features.

[0248] Decisions or rehabilitation plans at different levels are set by the administrator.

[0249] For the specific interaction process, please understand it in combination with the above method steps, and it will not be elaborated here.

[0250] Figure 3 is a schematic structural diagram of an electronic device provided by an embodiment of the present invention, as Figure 3 shown. Optionally, the electronic device 410 may include a first processor 2001.

[0251] Optionally, the electronic device 410 may further include a memory 2002 and a transceiver 2003.

[0252] Among them, the first processor 2001, the memory 2002, and the transceiver 2003 can be connected through a communication bus, for example.

[0253] The following will Figure 3 introduce each component of the electronic device 410 in detail:

[0254] Among them, the first processor 2001 is the control center of the electronic device 410, which can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 is one or more central processing units (CPUs), or can be an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. For example: one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs).

[0255] Optionally, the first processor 2001 can execute various functions of the electronic device 410 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0256] In a specific implementation, as an embodiment, the first processor 2001 may include one or more CPUs, such as Figure 3 CPU0 and CPU1 shown in

[0257] In a specific implementation, as an embodiment, the electronic device 410 may also include multiple processors, such as Figure 3 the first processor 2001 and the second processor 2004 shown in

[0258] Among them, the memory 2002 is used to store software programs for implementing the solution of the present invention and is controlled by the first processor 2001 for execution. The specific implementation method can refer to the above method embodiments and will not be elaborated here.

[0259] Optionally, the memory 2002 can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 2002 can be integrated with the first processor 2001 or exist independently and be coupled to the first processor 2001 through the interface circuit of the electronic device 410 ( Figure 3 not shown in the figure), and the embodiments of the present invention do not make specific limitations on this.

[0260] The transceiver 2003 is used to communicate with a network device or with a terminal device.

[0261] Optionally, the transceiver 2003 can include a receiver and a transmitter ( Figure 3 not shown separately in the figure). Among them, the receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0262] Optionally, the transceiver 2003 can be integrated with the first processor 2001 or exist independently and be coupled to the first processor 2001 through the interface circuit of the electronic device 410 ( Figure 3 not shown in the figure), and the embodiments of the present invention do not make specific limitations on this.

[0263] It should be noted that Figure 3 the structure of the electronic device 410 shown in the figure does not constitute a limitation on the router. The actual knowledge structure recognition device may include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0264] In addition, the technical effects of the electronic device 410 can refer to the technical effects of the gait analysis-based fall prediction and alarm method described in the above method embodiments, and will not be elaborated here.

[0265] It should be understood that the first processor 2001 in the embodiments of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0266] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM) are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0267] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0268] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context.

[0269] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0270] It should be understood that in various embodiments of the present invention, the magnitudes of the sequence numbers of the above processes do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0271] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0272] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described devices, apparatuses, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0273] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0274] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0275] In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0276] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0277] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A fall prediction and alarm method based on gait analysis, characterized in that, The method includes: S1. Collect multi-dimensional gait data vectors G of the foot, ankle, and knee joint parts of a rehabilitation patient respectively through electrode sheet sensors: G = [g1, g2, g3,..., g n , where g n represents the gait vector data collected by the nth local electrode patch sensor, including: the dynamic force line vector data of the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint part; S2. Perform feature processing on the multi-dimensional gait data vector G to extract gait data vector features of each part; S3. Based on the historical gait analysis database, predict and configure the proportion weights of the gait data vector features of each part; S4. Based on the gait data vector features and the proportion weights, construct a gait feature anomaly index and a fall risk index; S5. Determine whether gait abnormality occurs according to the gait feature anomaly index: If it occurs, determine the corresponding risk level according to the fall risk index, and match and output a fall alarm strategy corresponding to the risk level; If not, repeat the above steps.

2. The gait analysis-based fall prediction and alarm method according to claim 1, wherein S2. Perform feature processing on the multi-dimensional gait data vector G to extract gait data vector features of each part, including: Analyze the multi-dimensional gait data vector G to respectively obtain the active force line vector data corresponding to the foot, the joint movement trajectory vector data of the ankle, and the muscle contraction vector data of the knee joint part; Extract the gait data vector features of the action force line vector data, the joint movement trajectory vector data, and the muscle contraction vector data respectively: F force , F joint , F muscle , where: Where: f i represents the foot pressure value collected by the i-th pressure electrode patch sensor, represents the average foot pressure, and k represents the number of pressure patch sensors; Where: θ(t) represents the joint angle collected by the ankle angle electrode patch sensor and varying with time t, and θ ref (t) represents the reference joint angle (the angle under normal gait), and T represents the gait cycle; F muscle = max(EMG(t)) - min(EMG(t)), Where: EMG(t) represents the muscle electrical signal changing with time t.

3. The gait analysis-based fall prediction and alarm method according to claim 2, wherein S3. Based on the historical gait analysis database, predict and configure the proportion weights of the gait data vector features of each part, including: Statistically analyze the historical gait data vector features of each part and mark the corresponding anomaly labels; Calculate the feature importance score I of the historical gait data vector features of each part i , where i represents the i-th feature; According to the feature importance score I i , calculate the weight w of the historical gait data vector features of each part i : Among them, n represents the total number of features, and I j represents the feature importance score of the j-th feature; Construct a historical gait analysis database composed of several groups of {historical gait data vector features of each part, weight w i}, and use each group of data in the historical gait analysis database as the training input of the RF random forest model to perform model training and learning, and construct a weight prediction model; Verify the model performance of the weight prediction model. If the verification is passed, deploy and apply the weight prediction model; Predict and output the proportion weight w of the gait data vector features of each part by using the weight prediction model i .

4. The gait analysis-based fall prediction and alarm method according to claim 3, wherein S4. Based on the gait data vector features and the proportion weights, construct a gait feature anomaly index and a fall risk index, including: Construct the gait feature anomaly index S abnormal : S abnormal = w1·F force + w2·F joint + w3·F muscle , Where: w1, w2, and w3 are the weights of the action force line, joint movement trajectory, and muscle contraction characteristics respectively, and are calculated according to the weight w i and obtained by calculation; When S abnormal > the threshold value A, it is determined that a gait abnormality has occurred; Construct the fall risk index R fall : Where: α and β are respectively the weights of the change rates of the active force line and the joint movement trajectory, When Rfall > the threshold value B1, it is determined as a high fall risk, and there is a corresponding fall alarm strategy at this time; When the threshold B1 > R fall > the threshold B2, it is determined to be a medium fall risk, and there is a corresponding fall alarm strategy at this time; When the threshold B2 > R fall At this time, it is determined to be a low fall risk, and the fall alarm strategy is not output.

5. A fall prediction and alarm device based on gait analysis, the fall prediction and alarm device based on gait analysis is used to implement the fall prediction and alarm method based on gait analysis according to any one of claims 1-4, characterized in that, The device includes: A gait analysis terminal, including: MCU; Electrode sheet sensors, used to collect multi-dimensional gait data vectors G of the foot, ankle, and knee joint parts of a rehabilitation patient respectively; A signal processing module, used to perform feature processing on the multi-dimensional gait data vector G to extract gait data vector features of each part; A wireless transmission module, used to upload the gait data vector features to the background server; A power supply, used for power supply; The background server is used to predict and configure the proportion weights of the gait data vector features of each part based on the historical gait analysis database; and, based on the gait data vector features and the proportion weights, construct a gait feature anomaly index and a fall risk index; determine whether gait abnormality occurs according to the gait feature anomaly index: if it occurs, determine the corresponding risk level according to the fall risk index, and match and output a fall alarm strategy corresponding to the risk level; if not, repeat the above steps; The electrode sheet sensors, the signal processing module, the wireless transmission module, and the power supply are respectively electrically connected to the MCU; The gait analysis terminal is communicatively connected to the background server.

6. An electronic device, characterized in that, The electronic device includes: A processor; A memory, on which computer-readable instructions are stored. When the computer-readable instructions are executed by the processor, the method described in any one of claims 1 to 4 is implemented.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program code, and the program code can be called by a processor to execute the method according to any one of claims 1 to 4.

Images